Predicting the Rate Structure of an Evolved Metabolic Network

When glucose molecules are metabolized by a biological cell, the molecules are constrained to flow along distinct, reaction trajectories, which are defined by the cell’s underlying metabolic network. Using the computational technique of Elementary Mode Analysis, the entire set of all possible tra-jectories can be enumerated, effectively allowing metabolism to be viewed in a discretized space. With the resulting set of Elementary Flux Modes (EM), macroscopic fluxes, (of both mass and energy) that cross the cell envelope can be computed by a simple, linear combination of the individual EM trajectories. The challenge in this approach is that the usage probability of each EM is unknown. But, because the analytical framework we have adopted allows metabolism to be viewed in a dis-crete space, we can use the mathematics of statistical thermodynamics to derive the usage proba-bilities when the system entropy is maximized. The resulting probabilities, which obey a Boltz-mann-type distribution, predict a rate structure for the metabolic network that is in remarkable agreement with experimentally measured rates of adaptively evolved E. coli strains. Thus, in principle, the intracellular dynamic properties of such bacteria can be predicted, using only the knowledge of the DNA sequence, to reconstruct the metabolic reaction network, and the meas-urement of the specific glucose uptake rate.